This study proposes a comparison of the relative and combined efficacy of a narcotic antagonist (naltrexone) and behavior therapy in the treatment of narcotic addiction. Sixty heroin addicts, aged 18-25, from Ventura County will be randomly assigned to one of three treatment conditions: 1) Narcotic antagonist (naltrexone) alone, 2) behavior therapy alone, and 3) narcotic antagonist and behavior therapy. Each treatment group will be assessed on several measures: a) urine assays for presence of heroin and other drugs, b) therapy attendance, c) self-reported drug cravings, d) physiological response to videotapes of a person "shooting-up", e) job performance, f) assertion behavior (measured both by paper and pencil and role playing procedures), g) report of families and friends, and h) subjective measures of response to narcotic antagonists. Follow-up will range from a minimum of four months for some Ss to two years. In addition to the main question of comparative outcome across treatment conditions, several process issues will be addressed: a) an examination of the efficacy of assertion training with addicts, b) an examination of subjective reaction to antagonist induction compared to placebo, c) an examination of the efficacy of contingency contracting, d) an examination of electric shock and covert sensitization in the suppression of drug urges, e) a comparison of the effects of aversion therapy and extinction on drug cravings, and f) a comparison of the extinction of drug cravings between this project using narcotic antagonists and another Ventura County project using methadone.